Title:
Fighting Concurrency Bugs

Abstract:
Driven by the hardware shift to multicore architectures, concurrency is 
being brought into mainstream software development. Unfortunately, 
concurrent programs are prone to concurrency bugs, because of the 
inherent complexity of concurrency and the sequential thinking habits 
of programmers. Concurrency bugs' non-deterministic property also 
brings a lot of trouble to developers. Improving the reliability of 
concurrent programs is a critical and urgent task.

In this talk, I present two techniques to detect two types of concurrency 
bugs, both of which are important and not well-addressed before. 
These two techniques look at the concurrency bug problem from the 
angel of programmers' synchronization intention. The first one is 
AVIO, a technique to automatically infer programmers' atomicity 
intention and detect atomicity violation bugs. The second is MUVI,
 a technique to automatically infer variable correlation and
 detect multi-variable concurrency bugs. AVIO and MUVI effectively
 detect concurrency bugs, including some previously unknown 
bugs, from large concurrent systems. I will also briefly talk 
about how our characteristic study of real-world concurrency 
bugs motivated the above bug detection work and how to help 
expose concurrency bugs to support dynamic concurrency bug 
detection.